Computer exercise

A consumer study was conducted to understand consumer preferences for chocolate creams. For this purpose, 86 consumers were asked to evaluate 9 chocolate creams by putting scores between 0 (the cream is not liked) to 10 (the cream is liked a lot).

The data table has the consumers in rows, and in columns the 9 creams, and each entry gives the rating attributed by consumer i for cream k. So that differences in the use of the consumer rating scale do not play a decisive role in the analysis, the data have been centered for each consumer (centering by row, therefore).

Moreover, there are 3 variables describing consumers: age, sex and frequency of eating chocolate creams.

The data can be found in this file.

We want to split up the consumers into groups, that is, build homogeneous classes of consumers with similar preferences. Then we want to characterize these classes of consumers in terms of their descriptive variables (age, sex, frequency of eating chocolate creams).

Construct a clustering of the consumers by considering that the differences between consumers are due only to differences in appreciation of chocolate creams. For this clustering, you should keep the number of dimensions necessary to retrieve at least 90 % of the information contained in the data set.

Q1) If you keep the number of classes provided by default, how many are there?
2
3
4
5

Q2) Which qualitative variable best characterizes the clustering obtained?
age
sex
frequency

Q3) From the clustering results, which age group prefers Sveltesse chocolate creams?
under 25 years
25 - 40 years
+ 40 years

Q4) Which chocolate cream is most appreciated by the youngest age group?
Laitiere
Sveltesse
Nesquick
DanetteN
MtBlancN

Q5) What is the proportion of men in the cluster with the youngest members?
29 %
54 %
72 %
85 %
95 %

Q6) Please go ahead and analyze this data set more in detail, by interpreting the factor analysis results and the clustering results further.

Score =
Correct answers: